• State TI Thinks Resolved
  • Locked Locked
  • Replies 13 replies
  • Answers 3 answers
  • Subscribers 63 subscribers
  • Views 843 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TPS92515HV: PWM frequency and UVLO

Daniele Valente
Prodigy 140 points
Part Number: TPS92515HV
Other Parts Discussed in Thread: LM3409

Hello,

we're experimenting an issue with 10KHz PWM signal: if it goes to 0% we are not able to get a good output switching signal any more.

The datasheet says (www.ti.com/.../tps92515-q1.pdf, page 19):

"[...] if the PWM ON-time is shorter than the converter natural demanded ON-time, the PWM signal itself becomes the control signal for the high-side switch. The PWM pin activates a weak pulldown, as shown in Figure 25. Because the PWM pin is also UVLO (undervoltage lockout and device enable), when pulled low it is necessary to ensure the output is 100% OFF"

In our application there's a a spurious signal (that cannot be removed) on SW pin that come from the load. We observe that when the spurious signal goes over 8V with PWM duty cycle 0% something happens: we suppose that the "weak pulldown" is turned on, in this specific condition the regulator doesn't work, even if the PWM duty cycle goes to 100%.

The regulator start working normally only if we reduce the spurious signal (<8V).
Obviously during normal activity we cannot control the spurious signal.

Is this behaviour expected? How can we solve this issue?

Best Regards,
Daniele

  • 0
    TI__Mastermind 28655 points
    Hello,

    The SW needs to go to ground to refresh the boot capacitor. The idea of normal operation is the inductor has enough energy to pull the SW node load refreshing the boot capacitor otherwise the weak pull-down will allow it to restart. If it is a time issue a larger boot capacitor may work. A bootstrap charge circuit can work also but that will require more parts.

    LM3409 would not require this since it is a P-fet drive.

    Best Regards,
  • 0 in reply to Irwin Nederbragt
    Prodigy 140 points

    Hello,

    Thank you for your prompt reply. Can you suggest a bootstrap circuit that may work in our case?

    Thanks in advance,
    Daniele V.

  • 0 in reply to Daniele Valente
    TI__Mastermind 28655 points
    Hello Daniele,

    An isolated supply that can provide the correct voltage to the boot without loading the switch node (this is not trivial). It can also be done with a square wave of the correct amplitude and a small capacitor that will not add too much loading to the switch node, it will become switching loss (this will take some design work to function correct).

    Another way to do this would be to add a diode in series with your load to allow the pull-down to function.

    A simple way, if Vin is high enough, would be to add a diode and resistor (a resistor by itself may work), tied to boot from Vin to keep the boot capacitor charged when PWM is low. This only works if Vin is high enough compared to the load voltage. A small load may be needed to prevent this current from lighting the LED load. A clamp may need to be added to the boot as well.

    What is Vin? Vout? (entire operating range). How long is PWM low? (worse case).

    Best Regards,
  • 0 in reply to Irwin Nederbragt
    Prodigy 140 points

    Hello Irwin,

    Vin is 48V, Vout (if the output is enabled) between 25V and 35V, PWM signal can stay 0 for a while.

    Thanks for your help,
    Daniele

  • 0 in reply to Daniele Valente
    TI__Mastermind 28655 points
    Hello Daniele,

    That is enough headroom to provide some boot current from Vin with a high value resistor and possibly a clamp depending on the resistor value., it will effect your efficiency some and it has to be designed correct to prevent circuit failure. This would be the easiest method but may not be the best for what you are doing.

    Best Regards,
  • 0 in reply to Irwin Nederbragt
    TI__Mastermind 28655 points
    Hello Daniele,

    I haven’t heard back from you, I’m assuming you were able to resolve your issue.
    If not, just post a reply below (or create a new thread if the thread has locked due to time-out)

    Best Regards,
  • 0 in reply to Irwin Nederbragt
    Prodigy 140 points

    Hello,

    sorry for late answer, can you point me the circuit that you suggested in the last answer?

    Thanks and Regards,
    Daniele

  • 0 in reply to Daniele Valente
    TI__Mastermind 28655 points

    Hello Daniele,

    Please let me know the worst case Vin (highest value) and also the closest Vin and Vled can be.

    Best Regards,

  • 0 in reply to Irwin Nederbragt
    Prodigy 140 points

    Hello Irwin,

    Vin is 48V, let's say that minimum acceptable value is 44V while maximum is 50V. Maximum Vled is 36V, minimum 18V.

    Thanks,
    Daniele

  • 0 in reply to Daniele Valente
    TI__Mastermind 28655 points
    Hello Daniele,

    Try using a low leakage zener. A BZX84 at 4.7V with a 499 ohm series resistor, this is to prevent too much current from VCC when charging boot. Then add a 49.9K from Vin to boot and see how this works. This would need to be analyzed for all of their operating conditions including temperature. There isn't much room for tolerances due to abs max and boot UVLO max. If this is when Vout is lower than 36V the resistor value from Vin to Boot. I used 44V - 36V to get 8V, then 3.5V for boot UVLO leaving 4.5V headroom. Boot current at 90 uA gives 50 Kohm.

    Best Regards,
  • 0 in reply to Irwin Nederbragt
    Prodigy 140 points

    Hello Irwin,

    actually we use one diode 1N4148 between BOOT and VCC.
    Then I have to:
    - remove this diode and place a 4.7V zener with a series 500ohm resistor between BOOT and VCC (zener cathode on BOOT pin)
    - add one 49.9Kohm resistor between Vin and BOOT

    is it correct?

    Thanks,
    Daniele

  • 0 in reply to Daniele Valente
    TI__Mastermind 28655 points
    Hello Daniele,

    Leave the diode from VCC to boot (leave the existing boot circuit in place).

    Add a 4.7V zener with a series 499 ohm resistor across boot.

    Add a 49.9 Kohm from Vin to Boot (maybe with a series diode as well depending on duty cycle(s) and how the circuit is used).

    I would try this and look at the signals to see if it can be optimized further. I don't know how they intend to use the design so it would be good to check if this is adequate.

    "Try using a low leakage zener. A BZX84 at 4.7V with a 499 ohm series resistor, this is to prevent too much current from VCC when charging boot. Then add a 49.9K from Vin to boot and see how this works. This would need to be analyzed for all of their operating conditions including temperature. There isn't much room for tolerances due to abs max and boot UVLO max. If this is when Vout is lower than 36V the resistor value from Vin to Boot. I used 44V - 36V to get 8V, then 3.5V for boot UVLO leaving 4.5V headroom. Boot current at 90 uA gives 50 Kohm."

    Best Regards,
  • 0 in reply to Irwin Nederbragt
    TI__Mastermind 28655 points
    Hello,

    I haven’t heard back from you, I’m assuming you were able to resolve your issue.
    If not, just post a reply below (or create a new thread if the thread has locked due to time-out)

    Best Regards,